[adona058]

I am attempting to set up my dual-core computer to use MPIRUN with a case that I am running.

I decomposed the file, ran MPIRUN without a problem, and the results are distributed within the Processor0 and Processor1 directories as expected. Note that I have run MANY parallel cases on a larger network cluster in the past, so I am familiar with the general process.

The problem I am encountering occurs when I attempt to use reconstructPar to reconstruct the results. The terminal output is as follows:

reconstructPar
/*---------------------------------------------------------------------------*\
| ========= | |
| \\ / F ield | OpenFOAM: The Open Source CFD Toolbox |
| \\ / O peration | Version: 1.5.x |
| \\ / A nd | Web: http://www.OpenFOAM.org |
| \\/ M anipulation | |
\*---------------------------------------------------------------------------*/
Exec : reconstructPar
Date : Jul 03 2009
Time : 11:02:57
Host : adona058-laptop
PID : 27442
Case : /home/adona058/OpenFOAM/adona058-1.5.x/run/taylorBubbleParallel
nProcs : 1

// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //
Create time

fileName::stripInvalid() called for invalid fileName commandtouse
For debug level (= 2) > 1 this is considered fatal
Aborted



I would appreciate it if someone could get back to me on why this is occuring. I can set the debug level to 0 to avoid it, but I would rather know that the program is executing correctly.

Thanks,

Adam

[mattijs]

you have an invalid named file somewhere. After stripping out the invalid characters it is 'commandtouse'.

[hansel]

Originally Posted by mattijs

you have an invalid named file somewhere. After stripping out the invalid characters it is 'commandtouse'.

I have the same problem and my file names aren't using any strange characters. Is there a length limit for the entire path and file name?

[olesen]

Originally Posted by hansel

I have the same problem and my file names aren't using any strange characters. Is there a length limit for the entire path and file name?

The only real length limit would be INT_MAX, which likely isn't a problem (your filesystem will likely have problems first).

Could you post which file name is getting stripped, so we can reproduce the problem. For reference, here is what src/OpenFOAM/primitives/strings/fileName/fileNameI.H states:

inline bool Foam::fileName::valid(char c)
{
    return
    (
        !isspace(c)
     && c != '"'   // string quote
     && c != '\''  // string quote
    );
}

Thus whitespace and quotes are the only things it could (should) be complaining about.

[hansel]

The filename it was working on was:
/storage/OpenFOAM_cases/lenz2_size_exp/len2_18in_tsr06_halftail

The error message would report the name but leave off the _halftail part.

Changing the name to /storage/OpenFOAM_cases/lenz2_size_exp/len2_18inht_tsr06 made the problem go away.

I'm pretty sure I didn't have any spaces in the name, but since I changed it I can't be 100% sure.

This happened with 1.5dev, so you may not want to bother with it.

[flowris]

Hansel, Olesen,


I also tried to run blockMesh on a very deeply nested set of files, and it got the same error. When i copied the folder to my desktop, I could run it.

[mattijs]

Is this 16x on linux? We have not been able to reproduce the problem - do you have an example?

[flowris]

Goeiemorgen Mattijs,

I tried to run the interFoam tutorial by Hassan Hemida. The files can be downloaded from http://www.tfd.chalmers.se/~hani/kurser/OS_CFD_2007/. I nested everything very deeply in my file structure , namely in:

Documents/OpenFOAM\ info/Chalmers/interFoam/fillBottle/lamFillBottle/

However, some other minor difficulties occurred with these files. I had to change some names of parameters like gamma to alpha and pd to p in the system files. I also had to add files like constant/turbulenceProperties and constant/g. Is this because the tutorial was written for an older version of interFoam? I am using OpenFOAM-1.6.


Joris

[olesen]

Originally Posted by flowris

I tried to run the interFoam tutorial by Hassan Hemida. The files can be downloaded from http://www.tfd.chalmers.se/~hani/kurser/OS_CFD_2007/. I nested everything very deeply in my file structure , namely in:

Documents/OpenFOAM\ info/Chalmers/interFoam/fillBottle/lamFillBottle/

Spaces in directory and file names are pure poison for OpenFOAM (and several other bits of Unix software). Try renaming to something without space/tab etc and see if that does the trick.

[taheri.atp]

hi
I have this problem too!
I can run blockMesh but when I want too see my mesh in ParaFOAM I get this answer:

fileName::stripInvalid() called for invalid fileName /home/ahmad/Desktop/tpfoch/system/fvSolution::solvers:U|k|epsilon)
For debug level (= 2) > 1 this is considered fatal

can you help me to solve this problem?

[nimasam]

Dear Ahmad

it return to files names usually, check that you have no extra space or unusual file name, another suggestion can be that you can copy your blockMeshDict into icoFoam/cavity , then run blockMesh and see whether it works or not

[taheri.atp]

Originally Posted by nimasam

Dear Ahmad

it return to files names usually, check that you have no extra space or unusual file name, another suggestion can be that you can copy your blockMeshDict into icoFoam/cavity , then run blockMesh and see whether it works or not

wow!
I copied blockMeshDict to cavity and it worked!!!!
what does it mean?!?!?!

[nimasam]

it means nothing, means you should create a new test case or find where the typo is,
as i said it usually returns to bad file name

[wyldckat]

Greetings to all!

Since this is a frequently asked question, I've added to the FAQ: http://openfoamwiki.net/index.php/FA..._invalid.22.3F

Best regards,
Bruno

[Yeru]

Hi foamers,

in order to get familiar with OF and especially groovyBC, I'm currently trying to get a case of 2 cyclically linked patches to work (according to this example in the wiki).

Here's where the relation to this thread comes in:
simpleFoam stops after having calculated u in the first timestep (after startTime) with this:

word::stripInvalid() called for word p_Rect0front
    For debug level (= 2) > 1 this is considered fatal

What's p_Rect0front?
There are <PatchName>_front and <PatchName>_back for every patch in the inlet and outlet.
In my 0/p-file, in order to use groovyBC, I defined:

boundaryField{
[...]
innerRect0_front
{
    type        groovyBC;
    variables       "p_Rect0back{innerRect0_back}=sum(p*mag(Sf()))/sum(mag(Sf()));    p_Rect0front=sum(p*mag(Sf()))/sum(mag(Sf()));    f_relax=0.5;";
    valueExpression "(p_Rect0front + f_relax*(p_Rect0back - p_Rect0front))";
    value        $internalField;
    }
    innerRect0_back
    {
        type            zeroGradient;
    }

The problem will not be any special characters or length of the variables, since the wiki example uses

p_int22

, so including numbers and an underscore.

Any ideas?

[wyldckat]

Greetings Dominik,

I've Googled just now for:

site:www.cfd-online.com stripInvalid groovyBC

and found this post by Bernhard from 2012: http://www.cfd-online.com/Forums/ope...tml#post371428 post #2

Essentially you cannot have spaces inside the variables expressions, otherwise they'll count as part of the variable names. This is essentially exemplified in the wiki page section you indicated, where not a single space is present in the variables expressions!

And if you're wondering why this is still a problem since 2012, it's probably because no one has taken the time to report this on the bug tracker for swak4Foam/groovyBC .

Best regards,
Bruno

[Yeru]

Dear Bruno,

thanks a lot! It's sufficiently working now.
I'm sorry for stirring this up when a simple "google it" is the answer.
Anyway, thanks again,

Dominik

[bmayurm]

I am using command fluentMeshToFoam to convert .msh fluent mesh (from ICEM) to openfoam format.
The error is same above
for debug level (= 2) 1 this is considered fatal aborted (core dumped).
Please help. I'm New in this

[wyldckat]

Originally Posted by bmayurm

Please help. I'm New in this

Not enough information. Please follow the instructions given on this thread: http://www.cfd-online.com/Forums/ope...-get-help.html

[AJAY BHANDARI]

HI all,

I think my post fits best here . I am getting the same error as above. First my case ran correctly without any mistake and then when i compiled it again its giving error..

fileName::stripInvalid() called for invalid fileName 0
For debug level (= 2) > 1 this is considered fatal
Aborted (core dumped)

Please suggest what should i do?? Any help will be appreciated..

Regards
Ajay